Exclusionary abuses of dominance - the European Commission’s enforcement priorities

Fordham University Symposium
New York, 25^th September 2008

Introduction

When I planned my visit to Fordham some time ago I decided that it would be a good opportunity to speak on the European Commission's approach to unilateral exclusionary conduct of dominant firms. I thought that updating you on the latest steps of the review of our enforcement policy would be interesting and topical. I did not imagine at that time just how topical.

The recently published report of the DOJ on single firm conduct under section 2 of the Sherman Act, and the reaction from the FTC certainly make these times interesting. I believe that this report, together with the Commission's own forthcoming document, provide a vital opportunity to debate serious issues underlying enforcement policy in unilateral conduct. I hope today's discussions contribute to that debate.

Past and future steps

To start with my plans, I want the Commission to adopt a document that will guide our enforcement in applying Article 82 of the EC Treaty to exclusionary conduct by dominant firms. As a public administration, our resources are limited, and we have to make sure that our enforcement policy guides us to intervene where it matters most for business and consumers. We want to concentrate our action on cases where our action can be most effective.

Just before the summer, we sent a draft document setting out the Commission's enforcement priorities in applying Article 82 to the Member States – to the national competition authorities and responsible ministries. Earlier this month, we had a first discussion within the European Competition Network. Next week, this discussion will continue in a meeting with the Directors General of the various national competition authorities. Once the necessary discussions and consultations have finished and final amendments made, I want to submit this document for adoption to the College of Commissioners.

The content

Let me now turn to the main issues and give you an indication of the main conclusions we have drawn.

We all agree that the overall aim of competition policy, including the application of Article 82 and Section 2 of the Sherman Act, is to protect consumer welfare.

We also agree that protecting consumers is not necessarily the same as protecting competitors. We want to leave sufficient room for dominant firms to compete on the merits. Our enforcement should not protect competitors who do not deliver to consumers in terms of price, quality and innovation. If competitors leave the market because they are not good enough, that is the fair price of fair competition.

We also agree that distinguishing between competition on the merits and anticompetitive foreclosure may be particularly difficult for pricing conduct, as it is often hard to draw the line between fierce competition with low prices and pricing conduct that is likely to harm consumers. In this context I believe that it can be useful as a first step to examine on the basis of available cost and sales prices whether a hypothetical competitor as efficient as the dominant undertaking could meet the prices set by the dominant firm – the "as efficient competitor" test. If the data clearly shows that prices can be met by an equally efficient competitor, the Commission should usually not intervene - the competitive process would not be in danger and vigorous price competition would only be beneficial to consumers under such circumstances. However, one can already see some differences of opinion on the importance that should be given to this first screen in the overall assessment.

We also agree, of course, that we should in general apply an effects-based approach. We have already done this for a long time in our analysis of restrictive agreements under Article 81 and in merger control. An effects-based approach has also been central to our Article 82 enforcement in recent years. This has meant not only achieving greater consistency between our competition instruments in the EU, but also enhanced convergence with the US practice.

An effects-based analysis will not always require technical economic reasoning and evidence. Econometrics, for example, are a useful servant, but a terrible master. It is not hard to think of examples in which one could reach a conclusion of likely negative effects without venturing into a very elaborate exercise of data collection and identification of counterfactuals. For example, in the recent Tomra case, the dominant firm contractually prevented its customers from even testing its competitor’s products. No sophisticated economic analyses were needed to conclude that such conduct was liable to cause anticompetitive foreclosure.

Our point of view is that applying an effects-based approach does not require evidence of actual foreclosure; the likelihood of harmful effects is sufficient. Where actual harm has occurred, it of course strengthens the case for enforcement action.

Likely effects can sometimes be deduced from internal documents showing an exclusionary strategy by the dominant company. And in the absence of direct evidence, we will rely on broader evidence and our understanding of how markets function to assess whether the conduct is likely to have an exclusionary effect. We will take into account the capacity of competitors, customers and suppliers to counter the conduct of the dominant firm and the extent to which the conduct is applied in the market. In basic terms, we will look to see if there is a reasonable and convincing story of harm which explains why and how consumers are likely to be worse off as a consequence of the conduct.

We will not wait until actual effects have manifested themselves. If we wait until rivals are forced to leave the market then we have two serious problems.

First, you cannot resuscitate a corpse. No matter how effective the regulatory intervention, if it only happens after exit has occurred, then the damage to the market may be permanent.

Second, such intervention will completely miss many examples of consumer harm that weaken competitors, but do not kill them. Competitors may be wounded, confined to a small corner of the market, but not killed. Leaving these cases to one side is a recipe for serious under-enforcement.

Let me now turn to the role of efficiencies. In 2005, I shared with you my thoughts that the Commission should be open to efficiency defences, in the enforcement of Article 82 as in other areas. I suggested that such a defence should be structured like the efficiency defence in Article 81. Just as we apply the same effects-based approach across Article 81, 82 and the Merger Regulation, so should we apply the same approach to efficiencies. I remain firmly of that opinion today.

It would make little sense if the same conduct by the same company, for example an exclusivity agreement, is permitted under Article 81 but infringes Article 82, only because the text of Article 82 does not explicitly refer to efficiencies.

Taking this approach would mean that the dominant firm would have to demonstrate that the following cumulative conditions are fulfilled:

the efficiencies have been or are likely to be realised and are a result of the conduct;
the conduct is indispensable to the realisation of the efficiencies in the sense that there must be no realistic and less anti-competitive alternatives to the conduct that are capable of generating the same efficiencies;
the firm's conduct benefits consumers by producing efficiencies that outweigh any anticompetitive effects on consumers; and
the conduct does not eliminate effective competition, by removing all or most existing sources of actual or potential competition.

This as you will recognise means that we want to apply the 'consumer welfare balancing test' to unilateral conduct. This seems to us to be the right test for several reasons. First, it is the same test as we use for agreements and for mergers. Once again, it would be odd to adopt different approaches in these different areas. Second, it has broad support in the economic literature. Third, and most important, it protects consumers against conduct that, on balance, will harm them. That, after all, is the point.

What we try to do, again and again, is to find practical answers to the question: is this particular conduct likely to harm or benefit consumers? This is primarily an empirical question, not an ideological one; we should go where the evidence leads. If we take seriously our commitment to protect consumer welfare, then our policy choices and decisions should be driven by factual evidence and reasonable stories of harm, not ideology.

This is not to say that the exercise is easy. Look at technical tying, for example. Integrating distinct products into one new combined product may well benefit consumers, and we have to look very carefully at those benefits. But that does not mean that we should simply close our eyes to the possibilities for an incumbent firm with a strong dominant position to use technological tying to foreclose rivals with attractive alternative products from the market. The balancing test works perfectly well here as well.

Concluding remarks

Let me conclude.

It is clear that there are a range of views on these issues. In pursuing an enforcement policy, I think that a public agency has a duty to remain in the mainstream, to take positions based on an analytical framework that is broadly respected. That is what we have tried to do.

Not everyone in the US will agree with us, but it has always been simplistic to present these issues as US versus Europe. The reality has always been more subtle.

Looking at the range of views in the US, however, it is easy to see a greater concern about over-enforcement, about false positives, than in Europe. In Europe, we are worried equally about over-enforcement and under-enforcement. Both false positives and false negatives harm consumers.

One possible reason for this difference of emphasis, as Chairman Kovacic has suggested, may well derive from the differences in private enforcement in both systems. The combination of strong private enforcement with mandatory trebling of damages and jury trials may have contributed to make the Courts in the US more inclined to narrow the application of the rules than in Europe.

As you know I aim to strengthen private enforcement in Europe, but in a cautious way. We are not, for instance, proposing treble damages, Trebling damages for complex infringements – paradoxically – risks under-enforcement. It is easy to think of complex cases which, on balance, are considered harmful and where the victim should be compensated, but where treble damages seem to give a windfall gain to the plaintiff and impose an excessive burden on the defendant. In these circumstances, it is natural for courts to try and draw the enforcement border at a point where a treble damages punishment seems justified. But that would permit a swathe of cases which are nevertheless harmful. Hence under-enforcement.

We want to keep strong enforcement in Europe, both at the national and Community level. The review of our enforcement approach and the guidance document that I want to deliver will lead to better enforcement of Article 82. Our enforcement should be more effective, more efficient and focus on those cases that clearly lead to anti competitive effects. Already now cases based exclusively on Article 82 make up one quarter of the Commission's anti-trust cases. If you add to that cases involving both Articles 81 & 82, between 1/3 and ½ of our recent cases have involved Article 82.

In simple terms, Article 82 is and will remain central to our enforcement.